| Index | Recent Threads | Unanswered Threads | Who's Active | Guidelines | Search |
| World Community Grid Forums
|
| No member browsing this thread |
|
Thread Status: Active Total posts in this thread: 11
|
|
| Author |
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
In the past several hours I have been receiving the following message on 3 of 7 BOINC clients (both Windows and Linux):
----------------------------------------"Suspending computation - no recent user activity". I have never seen this message before and I am wondering if this could be cause for concern. I am using WCG clients 6.2.28 for Windows and 6.2.15 for Linux except on one machine that is testing 6.10.17 Linux (which appears to be doing well so far IMO). The unfortunate outcome of the "suspension" is the affected clients are shut down for a period of time before an indefinite amount of time is reached (about 5 minutes so far) before the WU's are resumed. Perplexed to say the least. Edit: Marked Subject [Resolved} [Edit 1 times, last edit by Former Member at Nov 15, 2009 9:10:49 PM] |
||
|
|
gb009761
Master Cruncher Scotland Joined: Apr 6, 2005 Post Count: 3010 Status: Offline Project Badges:
|
I am presuming (scene as you're a long term user), that you've already checked your settings, and determined as to whether each machine is using the Server or Local preferences...
----------------------------------------![]() |
||
|
|
JmBoullier
Former Community Advisor Normandy - France Joined: Jan 26, 2007 Post Count: 3716 Status: Offline Project Badges:
|
"Suspending computation - no recent user activity". My guess is that you have entered some number in the following setting of the website profile(s):Stop work after computer is idle for: 0.00 minutes This parameter is meant to allow computers to go sleeping when they have no other work to do. For the versions I know of this setting is not available in the Local Preferences, so you have to look in the device profiles you are using. Cheers. Jean. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
I am presuming (scene as you're a long term user), that you've already checked your settings, and determined as to whether each machine is using the Server or Local preferences... Yes, I double checked local preferences by hitting clear on each machine affected. My guess is that you have entered some number in the following setting of the website profile(s): Stop work after computer is idle for: 0.00 minutes This parameter is meant to allow computers to go sleeping when they have no other work to do. For the versions I know of this setting is not available in the Local Preferences, so you have to look in the device profiles you are using. Oops, [egg on face]. Yes I did have a couple profiles set with 0.0 and that explains why only some machines were affected. Thanks Jean |
||
|
|
JmBoullier
Former Community Advisor Normandy - France Joined: Jan 26, 2007 Post Count: 3716 Status: Offline Project Badges:
|
Oops, [egg on face]. Yes I did have a couple profiles set with 0.0 and that explains why only some machines were affected. Thanks Jean I find your wording ambiguous: my understanding is that when this parameter is set to 0 it means that this feature is not active. Can you please confirm that the problem happened for those machines linked to a profile where this parameter was set to more than 0.Save an egg: there are so many parameters to fit everybody's needs that it is easy to misuse one. Cheers. Jean. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
I find your wording ambiguous: my understanding is that when this parameter is set to 0 it means that this feature is not active. Can you please confirm that the problem happened for those machines linked to a profile where this parameter was set to more than 0. Ok, I am not sure how to respond to your observation. I have studied every message log for each my machines thoroughly. I have double checked all WCG and boincstats (bam) profiles and I can assure you that no profile anywhere has the setting "Stop Work for after computer is idle for..." set to any other number than 0.0. In my previous response, I misunderstood what you were saying thinking maybe 0.0 was wrong from what I had thought but I did not subsequently change any profile from 0.0 to any other number. For back ground information, I did upgrade one host to 6.10.17 and it ran WGC only quite well for about 4 days. I then unsuspended GPUGrid (suspended for months due to a slow GPU) to see if the client worked okay with WCG and my new GPU card. Things ran very well for several days without problems. I did do some tinkering with GPUGrid profile settings that I thought would not interfere with WCG. Apparently I was wrong and it seems project "cross-talk" may be occurring and causing problems. To date I have found two fixes that appear to resolve this issue: upgrade all clients to 6.10.17 or leave all clients at WCG approved releases. Apparently, the newest release is maybe an alpha male type version and over powers prior versions. Beyond that I have no clue what is going on without further research. Edit: By leaving all clients at WCG approved releases, I mean uninstalling 6.10.17 in favor of the WCG approved release. [Edit 1 times, last edit by Former Member at Nov 9, 2009 4:43:16 AM] |
||
|
|
JmBoullier
Former Community Advisor Normandy - France Joined: Jan 26, 2007 Post Count: 3716 Status: Offline Project Badges:
|
OK, things look more clear to me now, and I hope I will make it clear for you too.
----------------------------------------I presume the profile at GPUGrid has the same setting and that you set it to something more than 0. And regarding host profiles the last one loaded by a client is the one which counts for all projects to which this client is attached, until another one is loaded later. If you want that your WCG profile be the one that counts, force a change in it, save it, and when the next request for work goes to WCG it will be downloaded and take precedence. If you want to expedite this process you may also force an Update for WCG from the Projects tab. No need to downgrade your BOINC version if it works fine and if you want to run GPUGrid too. Cheers. Jean. |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
And regarding host profiles the last one loaded by a client is the one which counts for all projects to which this client is attached, until another one is loaded later. If you want that your WCG profile be the one that counts, force a change in it, save it, and when the next request for work goes to WCG it will be downloaded and take precedence. If you want to expedite this process you may also force an Update for WCG from the Projects tab.. Ah, this was the part I wasn't sure about but makes sense to me now. Thanks for the explanation. I will leave 6.10.17 on that one machine to continue WU's with WCG and GPUGrid and keep the WCG approved versions on the rest of the hosts ![]() |
||
|
|
Ingleside
Veteran Cruncher Norway Joined: Nov 19, 2005 Post Count: 974 Status: Offline Project Badges:
|
Ah, this was the part I wasn't sure about but makes sense to me now. Thanks for the explanation. I will leave 6.10.17 on that one machine to continue WU's with WCG and GPUGrid and keep the WCG approved versions on the rest of the hosts ![]() On GPUGrid and other BOINC-projects, the "Computing preferences" (earlier called Global preferences) clearly states: "These apply to all BOINC projects in which you participate. On computers attached to multiple projects, the most recently modified preferences will be used." It doesn't say it, but as long as you're using the same email-address on the multiple projects, these preferences will propagate to all the projects. Unfortunately when it comes to WCG, for one thing there's no mentioning the computing preferences works for all projects. And, even worse, the WCG-specific and the computing preferences is jumbled-together on the same page, with no indication what is project-specific and what is common for all BOINC-projects. For WCG you have: Global project-specific Global project-specific Global project-specific Global ![]() ![]() "I make so many mistakes. But then just think of all the mistakes I don't make, although I might." |
||
|
|
Former Member
Cruncher Joined: May 22, 2018 Post Count: 0 Status: Offline |
All 7 clients (6 WCG only and 1 WCG/GPUGrid) using only WCG profiles have been running stable for several days now without further issues. Since this was the first time I seriously attempted to run multiple projects concurrently, the profile issues became confusing. I had thought web based settings would be unique to each project but thanks to the information and clarifications provided in this thread, things are running great and I now consider this issue resolved. Thanks to all for the responses and help.
|
||
|
|
|